Materials and Methods

Two Midwestern racetrack facilities (Prairie Meadows Racetrack, Altoona, Iowa, and Remington Park Racetrack, Oklahoma City, Okla) were chosen for this study on the basis of their interest in participating and willingness to provide complete injury records for racing Quarter Horses. Records from 2000 through 2011 from both tracks were used to obtain information on total Quarter Horse race starts, CMIs, and available information for each horse with a CMI and each race during which a CMI occurred. Additionally, video of each race in which a CMI occurred was obtained. Past performance history was obtained for each horse with a CMI and for matched controls from a national Quarter Horse race reporting system.^a Matched control horses were defined as 2 randomly selected horses from the same race that were selected by drawing 2 post positions from those filled in the race, excluding the post position of the horse suffering the CMI. In addition, each racetrack or its referring veterinary institution supplied available documentation on pathological findings detailing the type of musculoskeletal injury sustained for horses with CMIs. Some of the cases in this study (the Quarter Horses from Remington Park Racetrack [n = 17] and Prairie Meadows Racetrack [13] for the years 2000 through 2006) were included in a previous publication¹⁷ comparing the incidence and describing the features of CMIs for Quarter Horses and Thoroughbreds at 3 Midwestern racetracks.

In the present study, a CMI was defined as an injury necessitating euthanasia by the attending racetrack veterinarian either during or immediately following a race. Catastrophic musculoskeletal injuries that occurred during training hours and injuries for which horses were not euthanized were not included. Catastrophic musculoskeletal injuries included those involving muscle, bones, tendons, and ligaments. Racing fatalities due to cardiovascular problems, including exercise-induced pulmonary hemorrhage, were not included in this study. Horses must have started the race (left the race gate) before incurring the injury to be included in the study.

The following variables were recorded for each injured horse and the 2 matched control horses from each race in which a CMI occurred: horse information (name, owner, trainer, and jockey), date of injury, racetrack (Prairie Meadows Racetrack or Remington Park Racetrack), sex (sexually intact male, gelding, or female), and age when the CMI occurred (years). The race information was recorded and included date of the race, racetrack surface (dirt or turf), racetrack condition (fast, good, sloppy, or muddy for dirt; firm or good for turf), race distance in race with CMI and distance of previous race for the horse (yards), type of race for race in which the CMI occurred and for the previous race (maiden or maiden special weight, maiden claiming, allowance, claiming, derby or futurity trial, and stakes race or handicap), and whether it was a state-bred incentive race (yes vs no). Environmental factors including temperature, weather conditions (rain, fog, cloudy, clear, windy, heat advisory, or cold advisory), and wind direction at finish line (head, tail, cross, or no wind) were recorded. The purse, weight carried, and, when available, speed rating for up to the last 5 races entered were recorded along with lifetime earnings, lifetime starts, age at first start, and whether the final race was a change in class (yes vs no; up vs down). When applicable, any layoff of ≥ 30 days prior to the race (yes vs no) and time of layoff (days) were noted as well as the number of days since the last start. A change in trainer or whether claimed within the last 5 starts was recorded (yes vs no). The finish position (if the horse finished), total lifetime yards raced, days from first start to CMI, days since last work, and prerace medication (phenylbutazone, furosemide, or both) were recorded. Details of the CMI included limbs injured, type of injury (fracture or soft tissue), and specific anatomic structures injured.

Video analyses, all done by 1 author (ABC) who had racing experience as an exercise rider, were used to obtain the following information: post position and number of horses in the race, reluctance to load (yes vs no), behavior in the starting gate (eg, normal, abnormal including rearing, and excessive movement), how the horse broke from the gate (eg, poorly, cleanly, or bumped), whether the horse stumbled or was cut off (yes vs no), whether the horse was bumped out of the gate (yes vs no) and whether the horse precipitated or received the bump, the lead out of the gate (left vs right), normal lead change through turn if applicable (yes vs no), lead when injury occurred (left vs right), lead changes associated with injury (yes vs no), number of lead changes (≥ 3 vs < 3), and cross firing (yes vs no) were recorded unless determined unobservable on the video. The position in the race relative to the rail and relative to the leader (lengths) or clear trailer, distance covered prior to injury (yards), stride quality prior to injury (normal, weaving, bumping, or erratic stride), whip use (yes vs no), jockey activity (on whip or urging, no activity, or unseated with no whip), effort when injury occurred (fatigue or dropping, driving, or holding), outside distraction (yes vs no), blinkers (yes vs no), and flipping halter (yes vs no) were recorded from the video.

Statistical analysis—Data were entered into a spreadsheet^b and imported into a statistical software package^c for analysis. The overall CMI incidence per racetrack, per year was calculated by dividing the number of CMIs for each group by the total race starts for the group and multiplying by 1,000. An ANOVA model was used to analyze the number of race starts, with racetrack and year as explanatory variables. The regression equation was as follows:

View Expanded

where is α is the intercept, β_i is the racetrack, and γ_j is the year. Residual plots were checked to confirm the normal distribution assumption, and the goodness of fit of the model to the data was verified (R² = 0.707).

The number of CMIs was analyzed by use of logistic regression, with racetrack and year as explanatory variables. The regression equation was as follows:

View Expanded

where again α is the intercept, β_i is the racetrack, γ_j is the year, and p is the probability of incidence. Goodness of fit was verified on the basis of the χ²/degrees of freedom ratio (11.0759/12), which was close to 1. Descriptive data for horses with CMI are presented as median (range) or percentage for each categorical variable.

Analysis of potential risk factors was done by comparison of the following variables between horses with CMI and control horses: sex, age, change (compared with previous race) in distance, change in class, change in purse, change in weight carried, change in speed rating, lifetime earnings, lifetime starts, first-time starter, age at first start, any layoff of ≥ 30 days prior to race, number of days since the last start, change in trainer or claimed within the last 5 starts, finish position and time in race if finished, total lifetime yards raced, days from first start or recorded work to breakdown, prerace furosemide, reluctance to load in the starting gate, behavior in the gate, post position, whether the horse stumbled or was cut off, whether the horse was bumped out of the gate and whether the horse precipitated or received the bump, the lead out of the gate, normal lead change through the turn if applicable, ≥ 3 lead changes, cross firing, position relative to the rail and relative to the leader or clear trailer, whip use, jockey activity, fatigue and stride quality prior to injury, outside distraction, blinkers, and use of a flipping halter. For the matched case-control data, stratified Cochran-Mantel-Haenszel statistics were used to test for an association between occurrence of a CMI as the categorical stratification (CMI or control) and the potential risk factors. Values of P ≤ 0.05 were considered significant.

Results

CMI incidence—From 2000 through 2011, there were 61,797 total Quarter Horse race starts at the racetracks studied. There were 82 CMIs, and the overall incidence was 1.33 CMIs/1,000 race starts. At Prairie Meadows Racetrack, there were 19,734 race starts with 24 total CMIs, yielding an incidence of 1.22 CMIs/1,000 race starts. At Remington Park Racetrack, there were 42,063 race starts with 58 total CMIs, yielding an incidence of 1.39 CMIs/1,000 race starts. There was a significant (P < 0.001) difference in the number of race starts between racetracks, but no significant (P = 0.541) difference between the total number of starts for both tracks combined, each year. There were no significant differences in the total number of CMIs between tracks (P = 0.522) or years (P = 0.656). There was no significant effect of year (P = 0.665) or racetrack (P = 0.634) on CMI incidence.

Characteristics of horses with CMIs—Complete information was available for 67 of 82 Quarter Horses that had a CMI. There were 23 (34.3%) 2-year-olds, 20 (29.9%) 3-year-olds, 16 (23.9%) 4-year-olds, 5 (7.5%) 5-year-olds, and 1 (1.5%) each of 6-, 7-, and 8-year-olds. Data for both the CMI and matched control groups were summarized (Table 1).

All 67 CMIs occurred on a dirt racetrack surface. Sixty-one (91.0%) CMIs occurred on a fast racing dirt surface. Four (6.0%) tracks were considered sloppy, and 2 (3.0%) were considered muddy. Race distances in yards run by the horses with CMI (median, 350 yards; range, 100 to 870 yards) were as follows: 350 (n = 17 [25.4%]), 330 (15 [22.4%]), 870 (9 [13.4%]), 400 (8 [12.0%]), 440 (5 [7.5%]), 250 (5 [7.5%]), and 300 (4 [6.0%]); 1 (1.5%) race each was run at 100, 220, 550, and 660 yards. Twelve (17.9%) horses were running in state-bred incentive races.

The median temperature at race time was 21°C (70°F; range, 9° to 33°C [48 to 91°F]). Of the 67 races in which a CMI occurred, 44 (65.7%) were run in sunny or clear conditions, 17 (25.4%) in cloudy conditions, and 6 (9.0%) in rainy conditions. At the finish line, there was a headwind in 17 (25.4%) races, a tailwind in 7 (10.4%), a crosswind in 8 (11.9%), and no wind in 35 (52.2%). Horses with a CMI were entered in maiden or maiden special weight (16/67 [23.9%]), claiming (15/67 [22.4%]), maiden claiming (10/67 [14.9%]), allowance (10/67 [14.9%]), stakes and handicap (9/67 [13.4%]), and derby trial (7/67 [10.4%]) races. The median purse of the race in which the CMI occurred was $12,000 (range, $2,500 to $65,000), and the median weight carried was 56.4 kg (124 lb; range, 53.6 to 58.2 kg [118 to 128 lb]).

Catastrophic musculoskeletal injuries frequently involved the right forelimb (38/67 [56.7%]), and the left forelimb was the second most frequent (16/67 [23.9%]). Both forelimbs were involved in 7 (10.4%) cases, and there was 1 (1.5%) each of left and right rear limb injury. Four (6.0%) axial musculoskeletal injuries resulted in the appearance of both rear limbs failing on the racetrack. These injuries included a lumbar vertebral fracture, a lumbosacral injury, a pelvic fracture, and muscular avulsions from the pelvis.

Fractures accounted for the majority of the CMIs (62/67 [92.5%]). Other injuries (5/67 [7.50%]) were limited to tendon, ligament, or muscle. These included 2 complete superficial and deep digital flexor tendon ruptures with rupture of the metacarpophalangeal or metatarsophalangeal joint (fetlock joint) capsule and additional soft tissue structures, a rear fetlock joint luxation, a suspensory apparatus failure from a laceration at the base of the proximal sesamoid bones, and a torn iliopsoas muscle.

Descriptive data for the CMIs were summarized (Table 2). Injuries that resulted in disruption of the metacarpophalangeal joint (proximal sesamoid bone, suspensory ligament, superficial and deep digital flexor tendon rupture, and a combination of proximal sesamoid bone and third metacarpal bone fracture) accounted for 23 of the 67 (34.3%) injuries. Seven additional horses had a recorded third metacarpal bone fracture, some of which likely involved the fetlock joint but were not specifically recorded as such.

After 2007, the use of phenylbutazone at Remington Park Racetrack did not require declaration. Documentation was available for 40 of the 67 horses with a CMI that ran at Prairie Meadows Racetrack over the study period or at Remington Park Racetrack prior to 2007, and only 1 was not administered phenylbutazone prior to the race in which the CMI occurred. We were unable to confirm this, but we believe nearly every horse was administered phenylbutazone prior to the race. As such, no additional analysis of this variable was performed.

The median number of horses in each race was 9 (range, 5 to 12); 1 of the 67 (1.5%) CMIs occurred in a race with 5 horses, 3 of 67 (4.5%) in races with 6 horses, 4 of 67 (6%) in races with 7 horses, 11 of 67 (16.4%) in races with 8 horses, 18 of 67 (26.9%) in races with 9 horses, and 30 of 67 (44.8%) in races with 10 horses. The inside post position was filled in all 67 races, but there were no CMIs in horses starting from that position. The numbers of horses with a CMI that broke from each gate were as follows: first gate, 0 of 67 (0%); second gate, 6 of 67 (9.0%); third gate, 11 of 67 (16.4%); fourth gate, 6 of 67 (9.0%); fifth gate, 9 of 67 (13.4%); sixth gate, 14 of 66 (21.2%); seventh gate, 6 of 63 (9.5%); eighth gate, 6 of 59 (10.2%); ninth gate, 7 of 48 (14.5%); and tenth gate, 2 of 30 (6.7%).

Twenty-seven of the 67 (40.3%) horses switched leads within 3 strides of when the injury became evident on the video. Nine of the 67 (13.4%) horses appeared to be switching leads in an effort to get off of the failing limb. Of the 67 horses, the lead limb that the horse initially broke from the gate on was injured in 36 (53.7%) and was not injured in 25 (37.3%); in the remaining 6 (9.0%), the injury was a hind limb or axial musculoskeletal injury. Of the 67 horses with CMIs, 46 (68.7%) led with the right limb and 21 (31.3%) with the left limb. Fifty-four horses had a unilateral forelimb injury. Thirty-five of the 54 (64.8%) led with the right limb, with 6 of the 35 (17.1%) having a left forelimb injury and 29 (82.9%) having a right forelimb injury. Nineteen horses led with the left limb, with 10 having a left forelimb injury and 9 having a right forelimb injury. Of the 7 horses with bilateral forelimb injuries, 6 were on the right lead and 1 was on the left lead. In the group of 9 horses that ran 870 yards, 8 had the correct inside lead in the corner. Two were injured in the turn, and 1 of the 2 was on the incorrect lead.

Of the 67 horses with a CMI, 2 (3%) were leading the race, 10 (14.9%) were within 1 length, 10 (14.9%) were within 2 lengths, 6 (9%) were within 3 lengths, 6 (9%) were within 4 lengths, 16 (23.9%) were within 5 lengths, 9 (13.4%) were within 6 lengths, 2 (3%) were within 7 lengths, and 6 (9%) were ≥ 8 lengths from the lead when the CMI occurred. Irrespective of race distance, most of the horses (47/67 [70.1%]) were injured after (21/67 [31.3%]) or within 10 yards before (26/67 [38.8%]) the finish line. The locations of the remaining injuries were spread out over the distance of the races. Few horses (5/67 [7.5%]) were injured within 50 yards of the gate: 2 horses in 400-yard races and 1 horse each in 330-, 350-, and 660-yard races. Two of 67 (3%) horses appeared to have a distraction at the rail: 1 dove inside away from vehicle headlights that were turned on during the race, and 1 tried to jump a shadow on the racing surface.

Matched case-control data—The horses with a CMI had significantly (P = 0.047) fewer starts (median, 6 starts; range, 1 to 51 starts) than did control horses (median, 6 starts; range, 1 to 50 starts). The median race finish was 6 (range, 1 to 10) for the 47 horses with CMIs that completed the race; this value was significantly (P = 0.006) behind that of the matched controls (median, 4; range, 1 to 10). Significantly (P = 0.032) more horses with a CMI stumbled at the break (28/67 [41.8%]), compared with matched controls (37/134 [27.6%]).

Of the 7 horses running an 870-yard race that completed the corner, 5 changed leads normally coming out of the corner; this was significantly (P = 0.046) different from the 14 matched controls, which all changed leads normally. Significantly (P = 0.003) more horses with CMIs (16/67 [23.9%]) than matched control horses (11/134 [8.2%]) were trailing in the race. Similarly, the horses with CMIs were a median of 4 lengths (range, 0 to 8 lengths) off the lead, and the control group was a median of 3 lengths (range, 0 to 9 lengths) off the lead, which was a significant (P < 0.001) difference. Stride quality was considered normal in 37 of the 67 (55.2%) horses with CMIs and 113 of the 134 (84.3%) controls (P < 0.001). The distribution of the effort when injury occurred in the CMI group (34 fatigued, 17 driving, and 16 holding) was significantly (P = 0.002) different than that of the control group (34 fatigue, 56 driving, and 44 holding). A distraction at the rail occurred in 2 of the 67 (3%) horses with a CMI and 0 of the 134 (0%) matched control horses, which was significantly (P = 0.046) different between groups.

Discussion

The present study was designed to determine the incidence and anatomic location of CMIs in racing Quarter Horses at 2 Midwestern racetracks over an 11-year period (2000 through 2011). Furthermore, we hoped that, by randomly selecting 2 matched control horses that raced in the same races as horses with CMIs and comparing a number of race variables, we might determine potential risk factors for CMIs in this population of racing Quarter Horses. For the 61,797 starts from 2000 through 2011 at the racetracks studied, there was an overall incidence of 1.33 CMIs/1,000 starts. There were 82 CMIs, with complete data available for 67 horses included in the study and 134 matched control horses. The evaluation of CMIs in racing Quarter Horses in this study was limited to CMIs occurring in horses that left the starting gate during a sanctioned race. These data were used because they were consistent between tracks and over time. The present study included only musculoskeletal injuries that resulted in euthanasia and excluded other causes of sudden death such as cardiovascular failure. Sudden death in racehorses during a race for nonmusculoskeletal reasons is important but uncommon, with the reported risk ranging from 0.08 to 0.3/1,000 starts in flat racing Thorougbreds.^23,24 In this study, we did not include injuries that occurred during training nor injuries that occurred in the starting gate, so comparison of CMI incidence between studies must account for this. The definition of CMI for this study was that the horse must break from the gate; however, no CMIs occurred in the starting gate during the time-frame of this study.

The incidence of CMIs in Quarter Horses in the present study was 1.33/1,000 race starts. A portion of this population (2000 to 2006) was included in a previous study.¹⁷ The incidence of CMIs in Quarter Horses at 3 racetracks in the prior study was 1.36/1,000 starts,¹⁷ and this remained essentially unchanged with the addition of 5 more years of data from the 2 tracks included in this study. The data in the present study were selected because they came from sources that could provide complete data for injuries and matched controls. There were no significant differences in CMI rates between tracks or over time. This could be viewed as positive; however, incidence could hopefully decrease with increased knowledge of risk factors, which could lead to changes in training and relevant modification in race conditions. The incidence reported for the present study is less than the overall incidence, including injuries during training, reported for Quarter Horses in California of 2.8 deaths/1,000 starts.¹⁸ When evaluating only CMIs occurring during a race, as for this study, the CMI incidence in California was 1.7/1,000 starts.¹⁸ At multiple racetracks in Texas, 0.8 catastrophic injuries/1,000 starts and 2.2 musculoskeletal injuries/1,000 race starts were reported.⁶ These data are different predominantly because a regulatory official determined whether it was likely that the horse could be saved. Our study included all horses that were euthanized, which may have included some horses that were euthanized on the basis of economic considerations. Overall, the data for the present study suggested that the incidence of CMIs in Quarter Horses in Midwestern racing jurisdictions compares favorably with that for horses racing in California and Texas. The incidence of CMIs in Quarter Horses was similar to that reported for Thoroughbred racing at the same tracks¹⁷ (1.48/1,000 starts) and in jurisdictions like California¹¹ (1.7/1,000 starts) and Kentucky¹⁴ (1.4/1,000 starts).

The horses with CMIs in the present study were predominantly 2- and 3-year-olds, which is a reflection of the predominant age of entries and races for these tracks.^d This is similar to racing Quarter Horses in Texas and California, where predominantly 2- and 3-year-olds developed CMIs.^6,21 The distribution of 37.3% females and 62.5% males in the present study was similar to the 67% of males developing a CMI in a previous study¹⁸ of 314 racing Quarter Horses. An association between sex and risk for CMI has not been found in Quarter Horses.⁶

In the present study, 25.4% of CMIs occurred in 350-yard races. The number of races at each distance was not included in our data, and it is likely that there were more races run at this distance. Similarly, many of the CMIs occurred in claiming races (37.3%); however, claiming races are common in both jurisdictions. It was not expected that only 4 of the 67 horses with a CMI had been claimed within the 5 previous starts. If horses with problems were being entered to have them claimed, more horses in the CMI group would be expected to have been claimed recently. We speculated that there may be more injuries in claiming races at these Midwestern tracks, considering that in the previous study¹⁷ from our group, it was reported that 40% of the CMIs in racing Quarter Horses in that study occurred in stakes-handicap races; however, this was not the case in the present study.

In racehorses, forelimbs are most commonly injured. In Thoroughbreds, some studies^11,14 indicate injury of the left forelimb is more common, and others^12,15 have shown no difference between forelimbs. In the only study¹⁷ that evaluated both Thoroughbred and Quarter Horses, the right forelimb was shown to be more frequently involved in Quarter Horses than Thoroughbreds. In a previous study,¹⁸ 81% of injuries in Quarter Horses were in a forelimb, with 63% in the right forelimb. This is similar to the finding of predominantly forelimb injuries (80.6%) in the present study, with 56.7% in the right forelimb. Specifically, there appeared to be a consistent predilection of right forelimb carpal and scapular injuries in the present study and in the previous study¹⁸ of racing Quarter Horses in California. In the present study, there was only 1 vertebral fracture. If we include the lumbosacral injury, the frequency of vertebral fractures (2/67 [29.9%]) would be less than in previous studies^6,18 (9% to 10%). The reason for the difference is not known. Vertebral and axial skeletal injuries manifested on the track as bilateral rear limb injuries that resulted in some of the more traumatic collapses seen in this study and others.²¹

When the injury occurred, 68.7% of the horses were leading with the right limb. Of the 67 horses, 52 (77.6%) horses leading with the right limb had a CMI of the lead limb (ie, the right forelimb or both forelimbs). This is similar to a video analysis that found 66% of horses with a forelimb fracture fractured the lead forelimb at the time of fracture,²⁵ and most limbs injured were lead forelimbs, irrespective of direction of racing.²⁶ In predominantly sprint races, horses are likely to lead with a preferred limb and will tend to use the same limb to lead.²⁷ Because 90% of Quarter Horses have a preference to lead with the right limb,²⁷ the frequency of injuries in the right limb in Quarter Horse racing is understandable. Peak vertical force is greatest in the lead forelimb, followed by the nonlead forelimb.²⁸ The maximal extension would occur with the right foot planted and the entire weight of the horse pivoting over the outstretched limb prior to a brief period of suspension.²⁹ It is well recognized that CMIs in racehorses are often the result of cumulative injury and failure of the bone during stress remodeling.^18,19,30 The consistency of racing Quarter Horses to lead with the right limb with maximum stress on the lead limb and the association of stress remodeling in CMIs would suggest that in racing Quarter Horses, prerace inspection should focus on the most commonly affected areas.

In this study of 67 horses with CMIs incurred during a race, the most common sites of injury were the carpus (35.8%) and the fetlock joint (34.3%), with third metacarpal bone fractures (10.5%) and scapular fractures (7.5%) less common. These data are similar between studies.^6,18 Injuries to the fetlock joint and carpus may be expected because the entire weight of the horse pivots over the maximally loaded joints. Scapular fractures, although uncommon in Thoroughbreds, occurred more frequently in Quarter Horses, predominantly in the right limb in this study and others.^18–20,30 The motion patterns of the shoulder have been evaluated in racing Quarter Horses and are specific to the leading and trailing limbs at a gallop.³¹ The lead limb has the greatest range of shoulder angle with a longer period of loading than the opposite limb.

In the present study, there was subjectivity in evaluation of the video and identification of when the injury occurred. It is possible that some injuries occurred earlier in a race but were not apparent. However, 70.1% of the horses showed clear evidence of the CMI within 10 yards of the finish line or after the finish line. This is similar to the 66% of injuries that were reported to occur after the finish line in Quarter Horse racing.²¹ A number of potential factors could explain the cluster of injuries near the finish line. Quarter Horses gain speed throughout the race and can have a maximum speed near 50 miles/h near the finish line.^21,32 If the horse is being urged by the rider, it can change the stride frequency and length.³³ At the finish line, there is an abrupt transition to the jockey coming up and slowing the horse and, at most tracks, as they go into a left turn. The potential for changes in loading pattern near maximum speed and load would occur near the finish line. Additional investigation into the kinematics of racing Quarter Horses at the finish line and during deceleration may provide a greater insight into these injuries and ways to reduce them.

For this study, use of the 2 matched controls from the same race provided a mechanism to attempt to identify the differences between horses that met selection criteria of being eligible and entered in the same races. A history of cumulative high-speed exercise has been identified as a risk factor for CMI in Thoroughbred racing¹¹ but was not identified in this study. Horses with a CMI had significantly fewer starts than the control horses. Other variables including total yards raced and days in training were not significant. Our data and previously published data⁶ do not support cumulative exercise as a risk factor in Quarter Horses, although further investigation is suggested.

Some examined variables occurred more frequently in the horses with CMI than the matched controls. Significantly (P = 0.032) more horses with CMIs (41.8%) stumbled at the break than the matched controls (27.6%). The stride quality was considered normal (55.2%) in significantly (P < 0.001) fewer of the CMI group than the matched control (84.3%). Significantly (P = 0.002) more horses with a CMI (50.1%) appeared fatigued or were dropping back when the injury occurred, compared with matched controls (25%). Significantly (P = 0.003) more horses with a CMI (25%) were trailing in the race when the injury occurred, compared with matched controls (8.2%). Similarly, Cohen et al⁶ reported stumbling during the race and whip usage were significant risk factors for an injury. In Quarter Horse sprint racing, despite these variables being evident on video, it would be difficult for jockeys to both recognize the situation and effectively slow a horse before catastrophic injury occurs.

These data further document some of the unique risk factors seen in Quarter Horse racing. Although the incidence of CMI is similar between Thoroughbreds and Quarter Horses, there are differences in the types of injuries and risk factors for CMIs during a race. We suggest that mechanisms to reduce such injuries in racing Quarter Horses and ongoing research efforts may need to focus on the right forelimb and actions that occur near the finish line.